diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..bb026ec
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+data/*.log
diff --git a/app.py b/app.py
index 0a78b18..e14663e 100644
--- a/app.py
+++ b/app.py
@@ -1,128 +1,136 @@
 """
 Copyright (C) 2018 Kunal Mehta <legoktm@member.fsf.org>
 
 This program is free software: you can redistribute it and/or modify
 it under the terms of the GNU Affero General Public License as published by
 the Free Software Foundation, either version 3 of the License, or
 (at your option) any later version.
 
 This program is distributed in the hope that it will be useful,
 but WITHOUT ANY WARRANTY; without even the implied warranty of
 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 GNU Affero General Public License for more details.
 
 You should have received a copy of the GNU Affero General Public License
 along with this program.  If not, see <https://www.gnu.org/licenses/>.
 """
 
 import elasticsearch
 import elasticsearch.client
 import functools
 import json
 import os
 from php import Php
 import re
 import requests
 import sys
 import toolforge
 
+__dir__ = os.path.dirname(__file__)
 JENKINS = 'https://integration.wikimedia.org/ci/'
 TRIGGERED_RE = re.compile('Triggered by change:\n    <a href="(.*?)">([0-9,]*?)</a>', flags=re.MULTILINE)
 CREDENTIALS = Php.parse_ini_file(os.path.expanduser('~/.elasticsearch.ini'))
 toolforge.set_user_agent('flaky-ci')
 session = requests.Session()
 INDEX = 'flaky-ci'
 DOC_TYPE = 'jenkins-job'
 client = elasticsearch.Elasticsearch(
     ['tools-elastic-01.tools.eqiad.wmflabs:80'],
     connection_class=elasticsearch.RequestsHttpConnection,
     http_auth=(CREDENTIALS['user'], CREDENTIALS['password'])
 )
 
 
 def jenkins_api(path):
     r = session.get(JENKINS + path + '/api/json')
     r.raise_for_status()
     return r.json()
 
 
 def build_log(job_name, number):
     r = session.get(JENKINS + 'job/%s/%s/consoleText' % (job_name, number))
     r.raise_for_status()
     return r.text
 
 
 def recent_builds(job_name):
     data = jenkins_api('job/%s' % job_name)
     latest = data['builds'][0]['number']
     first = data['firstBuild']['number']
     for i in reversed(range(first, latest + 1)):
         yield build_info(job_name, i)
 
 
 def init_indices():
     indices = elasticsearch.client.IndicesClient(client)
     if indices.exists(index=INDEX):
         indices.delete(index=INDEX)
     with open(os.path.join(os.path.dirname(__file__), 'index.json')) as f:
         index = json.load(f)
     indices.create(
         index=INDEX,
         body=index
     )
 
 
 @functools.lru_cache()
 def build_info(job_name, number):
     data = jenkins_api('job/%s/%s' % (job_name, number))
     document = {}
     if data.get('description'):
         triggered = TRIGGERED_RE.search(data['description'])
         if triggered:
             document['triggered'] = triggered.group(2)
     params = {}
     for param in data['actions'][1]['parameters']:
         params[param['name']] = param['value']
     document['project'] = params['ZUUL_PROJECT']
     document['job'] = job_name
     document['number'] = number
     document['result'] = data['result']
-    document['console'] = build_log(job_name, number)
+    job_dir = os.path.join(__dir__, 'data', job_name)
+    if not os.path.isdir(job_dir):
+        os.mkdir(job_dir)
+    # data/job/number-FAILED.log
+    log_fname = os.path.join(job_dir, '%s-%s.log') % (number, data['result'])
+    if not os.path.exists(log_fname):
+        log = build_log(job_name, number)
+        with open(log_fname, 'w') as f:
+            f.write(log)
     document['node'] = data['builtOn']
     # Unix times
     document['timestamp'] = data['timestamp'] / 1000
     document['duration'] = data['duration'] / 1000
 
     return document
 
 
 def import_to_es(document):
     print('Importing %s #%s' % (document['job'], document['number']))
-    print(len(document['console']))
     print(client.index(index=INDEX, doc_type=DOC_TYPE, body=document))
 
 
 def main():
     init_indices()
     jobs = ['mwext-mw-selenium-node-composer-jessie']
     for job_name in jobs:
         for document in recent_builds(job_name):
             import_to_es(document)
 
 
 def search():
     q = client.search(index=INDEX, body={
         'query': {
             'regexp': {
                 'console': 'An element could not be located on the page using the given search parameters',
             }
         }
     })
     print(q)
 
 
 if __name__ == '__main__':
     if '--import' in sys.argv:
         main()
     else:
         search()
diff --git a/data/.gitkeep b/data/.gitkeep
new file mode 100644
index 0000000..e69de29
diff --git a/index.json b/index.json
index 88495ea..f3c5408 100644
--- a/index.json
+++ b/index.json
@@ -1,60 +1,55 @@
 {
   "mappings": {
     "jenkins-job": {
       "dynamic_templates": [
         {
           "string_fields": {
             "match": "*",
             "mapping": {
               "index": "not_analyzed",
               "type": "string",
               "omit_norms": true
             },
             "match_mapping_type": "string"
           }
         }
       ],
       "_all": {
         "norms": false,
         "enabled": true
       },
       "properties": {
         "number": {
           "type": "long"
         },
         "triggered": {
           "norms": false,
           "type": "text"
         },
         "job": {
           "norms": false,
           "type": "text"
         },
-        "console": {
-          "type": "text",
-          "norms": false,
-          "analyzer": "keyword"
-        },
         "project": {
           "norms": false,
           "type": "text"
         },
         "duration": {
           "type": "float"
         },
         "timestamp": {
           "type": "float"
         },
         "result": {
           "norms": false,
           "type": "text"
         },
         "node": {
           "norms": false,
           "type": "text"
         }
       }
     }
   }
 }